Ex-IDF intel chief: AI, tech will not suddenly revolutionize war
New technologies like artificial intelligence “will neither cause a revolution on the battlefield in the next 10 to 15 years nor... ensure military victory for those who master it,” said former IDF Military Intelligence chief Amos Yadlin on Wednesday.
Yadlin’s analysis came in a position paper published as part of the ongoing World Economic Forum in Davos.
The intelligence chief and current Institute for National Security Studies executive director first noted that “the race to build next-generation technology is reshaping the contours of global relations.”
He described “the competition between the United States and China to become the world leader in artificial intelligence (AI), quantum computing and fifth-generation wireless networks” as capturing much of the world’s imagination.
Furthermore, he stated that many predict that “these technologies will bring about the next revolution in military affairs by improving the collection and analysis of data, encryption, the transfer of large amounts of information and the creation of autonomous weapons systems.”
Next, he refuted the obsession with new technologies in order to cause a complete and sudden military revolution, citing the “power of three twos” formulated by IDF Brig.-Gen. (res.) Isaac Gat.
Yadlin said that the first “two” is that the development and production of every new weapons system is “twice as expensive as initially estimated.” The second “two,” he said, was that innovation “takes twice as long as expected to build operationally significant capabilities.”
Regarding the final “two,” he explained that “new weapons systems tend to be only half as effective on the battlefield as predicted.”
Yadlin then listed off other issues that could limit the dominance of new technologies like AI. He said that, throughout history, anytime a new technology is introduced, it is eventually mitigated to some extent by countermeasures.
Humans may not be as fast as AI, but they are uniquely talented at probing for vulnerabilities.
For example, he said that AI-based systems might be fooled by an inability to read unpredictable situations, the way that small alterations to people’s appearances have helped fool expensive facial recognition technology.
Moreover, he said that “the logic of war can be the opposite of the logic of everyday life .... It may be especially difficult to train computers to understand when or how to act in a counterintuitive manner.”
Yadlin also addressed the fact that military actions themselves are only as useful as the properly defined goals they are designed to achieve. No technology can help a country succeed on the battlefield if the goals themselves are amorphous or unrealistic, stated Yadlin.
Finally, Yadlin said that many of the worst-case scenarios that keep decision-makers up at night are still far-off from actually taking place. He said that a “Pearl Harbor” style cyberattack is unlikely because most countries lack the required cyber capabilities, and those that have the ability will not use them against each other for fear of a massive cyber counterattack.